Technology with a Purpose

Ships produce lots of data

Today’s ships are equipped with numerous sensors and advanced systems which help the crew to operate vessels more efficiently. However, the data produced onboard is far from being utilized to its full extent. By tapping into this unexploited data stream, substantial fuel efficiencies are enabled.

We use science for insight

We collect 2.1 Billion measurements every day from hundreds of vessels – that’s over 24,000 measurements per second. Our database of operational vessel sensor data already includes 96,000 sea days of high accuracy real-time data.

Eniram rack server
Eniram data bank

We build a mathematical model of the vessel that surpasses the accuracy of other existing modelling methods. This enables the connection between crew operations and fuel consumption to be clearly seen and the impact of influencing factors such as current, wind, waves, trim, etc., to be precisely pinpointed and quantified. Previously this was not possible because these small changes in performance got buried under constantly changing external factors. One needs a method with extreme precision to remove these fluctuations and uncover the subtle changes enabling the savings.

Our approach to vessel performance optimization is unique. The method combines naval architectural knowledge and state-of-the-art machine-learning algorithms. It is based on real measured data, not utilizing any artificial counterparts such as performance curves from simulations or laboratory tests (including CFD simulations, wind tunnel tests, open water propeller curves, towing tank tests and manufacturer test bench curves). The fact is the reality never matches the laboratory.

Propulsion Power Decomposition is our key method



Process of Propulsion Power Decomposition

    1. Collect data from ship systems and meters (sample frequency: 1.16Hz ~ 4 million measurements / day)
    2. Combine this with a generic mathematical ship performance model developed by Eniram
    3. Get a statistical analysis, a regression, for a specific ship by accommodating the generic model with the data collected from a specific ship.

Eniram Developers’ Blog

Eniram Developers have established a tech blog. If you’re a techie, click here

Topics include Measuring geodetic distance calculation performance and Reactive programming in Python.Developers blog